2 December 2016

Airbnb in Berlin

  • 20,000+ Berliners hosted more than half a million guests in 2015
  • Renting out flats as a source of income has become more and more popular.
  • "Zweckentfremdungsverbot" law was passed in May 2014 but only came into force in April 2016.

Airbnb in Berlin


Source: www.insideairbnb.com

Evidence from Literature

  • Airbnb drives up rent: Schäfer et al. (2016) found that rent growth is higher in neighbourhoods that have a significant number of misused flats.
  • Effect on the hotel industry: Guttentag (2015) estimates that Airbnb sold about 15 million room nights in 2012.
  • Evidence from Texas: Zervas, Proserpio, and Byers (2016) found that a 10 percent size increase of the Airbnb market in Texas resulted in a .39 percent decrease in hotel revenue.

Research Question & Hypothesis

What is the effect of Airbnb on Hotels in Berlin?

The higher the Airbnb supply in a given district in Berlin, the greater the negative effect on the hotel industry in that same district.

Data Sources

Our paper uses data from various sources for the period of 2010 to 2014:

  • Statistical Information System Berlin/Brandenburg (SBB)

  • Federal Statistical Office and the statistical offices of the Länder (FSO)

  • InsideAirbnb.com

  • Eurostat

Main Variables

Dependent variable:

\[{Occupancy Rate}_{it} = \frac {{Overnight Stays}_{it}}{{Hotel Beds}_{it}*{days}_{t}}\]

Main independent variables:

  1. Airbnb Supply: Cumulative sum of new listings

  2. Dynamic Airbnb Supply: Active listings based on reviews

Airbnb Listings per Neighbourhood (static model) (2010 - 2014)

Airbnb Listings per Neighbourhood (dynamic model) (2010 - 2014)

Effect of Increase in Airbnb Listings on Berlin Hotel Occupancy Rates (2010 - 2014)

Regression Model

Fixed Effects (for time and district) to account for unobserved heterogeneity:

\({\log Occupancy Rate}_{it} = \beta _i * \log AbbSupply_{it} + \beta _j * X' + \tau _i + \varepsilon _{it}\)

\({X'}\): Control variables (UE rate, average HH income, air passengers arriving in Berlin)

\({\tau _i}\): Time and district dummies

Estimation Results

Dependent variable:
Occupancy Rate
LM (1) FE (2) FE (3) FE (4)
Log Airbnb Listings 0.009*** 0.026*** -0.010*** -0.013***
(0.003) (0.004) (0.002) (0.002)
Average HH Income (Log) -0.139* -0.388* -0.280*** -0.243***
(0.078) (0.202) (0.096) (0.094)
Unemployment Rate 0.141 0.862** 0.351* 0.345*
(0.152) (0.406) (0.194) (0.188)
Incoming Passengers 0.00000*** 0.00000***
(0.000) (0.000)
Market Entry 0.036***
(0.006)
Neighbourhood-specific trend Yes No No No
Time trend Yes No No No
District/Time FE? No Yes Yes Yes
Observations 720 720 720 720
R2 0.996 0.077 0.790 0.802
Adjusted R2 0.995 0.075 0.773 0.783
Note: p<0.1; p<0.05; p<0.01

Estimation Results (2)

Dependent variable:
Occupancy Rate
Airbnb Listings 0.00004**
(0.00002)
Sqrd. Airbnb Listings -0.000
(0.000)
Average HH Income (Log) -0.858***
(0.118)
Unemployment Rate 0.825***
(0.170)
Incoming Passengers 0.00000***
(0.000)
Market Entry 0.027***
(0.006)
District/Time FE? Yes
Observations 720
R2 0.795
Adjusted R2 0.775
Note: p<0.1; p<0.05; p<0.01

Preliminary Conclusion & Limitations

  • Our model suggests a significant inverse relationship between Log Airbnb supply and hotel occupancy rates
  • Our limited time period excludes a crucial year (2015).
  • The timing of the Zweckentfremdungsverbot law's effect is leads to skewed Airbnb supply in our time period.

Thank you!


Questions?

Further Reading

Guttentag, Daniel. 2015. “Airbnb: Disruptive Innovation and the Rise of an Informal Tourism Accommodation Sector.” Current Issues in Tourism 18 (12). Taylor & Francis: 1192–1217.

Schäfer, Philipp, Nicole Braun, Richard Reed, and Nicole Johnston. 2016. “Misuse Through Short-Term Rentals on the Berlin Housing Market.” International Journal of Housing Markets and Analysis 9 (2). Emerald Group Publishing Limited.

Zervas, Georgios, Davide Proserpio, and John Byers. 2016. “The Rise of the Sharing Economy: Estimating the Impact of Airbnb on the Hotel Industry.” Boston U. School of Management Research Paper, no. 2013-16.